Glossary术语表AdaptiveNeuro-FuzzyInferenceSystem(ANFIS)AtechniqueforautomaticallytuningSugeno-typeinferencesystemsbasedontrainingdata.Foreword（前言）Thepastfewyearshavewitnessedarapidgrowthinthenumberandvarietyofapplicationsoffuzzylogic.Theapplicationsrangefromconsumerproductssuchascameras,camcorders,washingmachines,andmicrowaveovenstoindustrialprocesscontrol,medicalinstrumentation,decision-supportsystems,andportfolioselection.过去几年间，模糊逻辑无论是在应用数量上还是应用种类上都呈现快速增长的趋势。其应用范围从消费产品，例如照相机，便携式摄像机，洗衣机，及微波炉到工业过程控制，医药器具，决策支持系统，以及部长职务选举等Tounderstandthereasonsforthegrowinguseoffuzzylogicitisnecessary,first,toclarifywhatismeantbyfuzzylogic.为了理解模糊逻辑为何能得以如此快速使用，首先，有必要理清什么是模糊逻辑。Fuzzylogichastwodifferentmeanings.Inanarrowsense,fuzzylogicisalogicalsystem,whichisanextensionofmultivaluedlogic.Butinawidersense,whichisinpredominantusetoday,fuzzylogic(FL)isalmostsynonymouswiththetheoryoffuzzysets,atheorywhichrelatestoclassesofobjectswithunsharpboundariesinwhichmembershipisamatterofdegree.Inthisperspective,fuzzylogicinitsnarrowsenseisabranchofFL.Whatisimportanttorecognizeisthat,eveninitsnarrowsense,theagendaoffuzzylogicisverydifferentbothinspiritandsubstancefromtheagendasoftraditionalmultivaluedlogicalsystems.模糊逻辑有两种含义。从狭义上来说，模糊逻辑是一个逻辑系统，它是一种多值逻辑的扩充。但是广义上来说，就今天多数使用情况来看，模糊逻辑（FL）几乎与模糊理论的同义，这里模糊理论是一种与事物无明显界限的的分类，在这个界限里关系是一个度的问题。以此观点，狭义上说，模糊逻辑应该是模糊逻辑理论（FL）的一个分支。重要的是，我们要认清楚模糊逻辑与以往传统的多值逻辑系统无论是在精神还是主旨上都不相同，甚至从狭义角度也不相同。IntheFuzzyLogicToolbox,fuzzylogicshouldbeinterpretedasFL,thatis,fuzzylogicinitswidesense.ThebasicideasunderlyingFLareexplainedveryclearlyandinsightfullyintheIntroduction.WhatmightbeaddedisthatthebasicconceptunderlyingFListhatofalinguisticvariable,thatis,avariablewhosevaluesarewordsratherthannumbers.Ineffect,muchofFLmaybeviewedasamethodologyforcomputingwithwordsratherthannumbers.Althoughwordsareinherentlylessprecisethannumbers,theiruseisclosertohumanintuition.Furthermore,computingwithwordsexploitsthetoleranceforimprecisionandtherebylowersthecostofsolution.在模糊逻辑工具箱，模糊逻辑应该理解为FL，即模糊逻辑的广义定义。在介绍中，FL的潜在的基本的意义已经解释的非常清楚和明白。这里，需要增加说明的是，FL潜在的基本含义是一种关于语言上的变量，也就是说，一种既有数字又有语句的变量。事实上，绝大多数，FL在计算机处理中作为方式逻辑通常被视为语句而不是数字。尽管在精度上，语句与生俱来就比数字低，但是其使用更接近人类的直觉认识。另外，语句的计算处理开拓了imprecision的容度，因此也减小了解决费用。AnotherbasicconceptinFL,whichplaysacentralroleinmostofitsapplications,isthatofafuzzyif-thenruleor,simply,fuzzyrule.Althoughrule-basedsystemshavealonghistoryofuseinAI,whatismissinginsuchsystemsisamachineryfordealingwithfuzzyconsequentsand/orfuzzyantecedents.Infuzzylogic,thismachineryisprovidedbywhatiscalledthecalculusoffuzzyrules.ThecalculusoffuzzyrulesservesasabasisforwhatmightbecalledtheFuzzyDependencyandCommandLanguage(FDCL).AlthoughFDCLisnotusedexplicitlyintheFuzzyLogicToolbox,itiseffectivelyoneofitsprincipalconstituents.Inthisconnection,whatisimportanttorecognizeisthatinmostoftheappli...
